On 12/11/06, Pete Robbins <[EMAIL PROTECTED]> wrote:
OK... I'm a little wary of the SDO changes! I think having a stack of the namespaces and writing them as required on the element as they are needed is exactly right, in fact that is how I wrote it originally ;-) Since then, however, some changes have removed this logic! You may have seen some of the old code commented out. Now there are 3 reasons why this may have been removed: 1) it didn't work (unlikely as I coded it ;-)) ) 2) it did not conform to the SDO specification 3) it was causing a problem for a user ... probably the SDO PHP project I'd put my money on 3) so I need to do some checking and find the change that removed this and try and understand why it was done.
Actually... I think the answer is 4) The idea being to add all the xmlns definitions on the root element! The parsing fix (TUSCANY-920) looks fine and, as you say, will work in 99% of cases. I think it only falls down if someone has defined the same prefix with different namespace uri on different branches of the tree in the document.... unlikely but can be fixed using a stack of namespace maps.
I'll go throught the 14 failing tests and see if there are any real problems there or it is just a matter of re-generating the test comparison files. Cheers, On 12/11/06, Jean-Sebastien Delfino <[EMAIL PROTECTED]> wrote: > > Pete Robbins wrote: > > Sebastien, > > > > Go ahead and commit the changes. I'll review them once committed. The > > logging changes sound good. > > > > I'm pretty sure the xsi:type writing has been fixed (many times!) so > I'd > > like to look at the latest change. > > > > Cheers and good luck! > > > > Pete, Thanks for helping review this. I've been able to merge with no > major conflict so I'm going to commit the SCA changes. > > A minor change to SDORuntimeException has been committed under revision > r473864. > > The other SDO changes to the XML writer and parser are isolated in two > patches attached to JIRAs TUSCANY-919 and TUSCANY-920. > > Thanks > > -- > Jean-Sebastien > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > -- Pete
-- Pete
